Indoor/Outdoor (IO) detection (IOD) using Wi-Fi- and smartphone-based technologies is in high demand and interest in both the industrial and research fields. This paper proposes a novel and effective hybrid IOD (HIOD) approach for detecting a smartphone user's IO location. The HIOD approach uses signals received from both Wi-Fi and GPS as well as the latest positioning technologies such as multilateration, fingerprinting and machine learning. This paper proposes and implements two-level signal-to-noise ratio (SNR) threshold parameters for the first time, which are specifically derived from GPS signals through 42 empirical tests at seven test sites with adequate environmental factors considered. Using the newly derived IOD threshold parameters and a set of IO detection rules, the HIOD approach is then tested at 20 test points (TPs) in a city canyon area, where most of the TPs are under semi-indoor or semi-outdoor conditions. The final test results show that a 100% IOD rate is achieved.
